As part of the National Cancer Institute's Cancer Genome Anatomy Project (CGAP), the Genetic Annotation Initiative (GAI) seeks to expand the collection of gene-based genetic analysis reagents for cancer research. We continue to identify and have now identified more than 40,000 high-probability candidate single nucleotide polymorphisms (SNPs) by analyzing publicly available expressed sequence tag chromatograms with a set of sequence analysis tools. This approach has also been applied to identify more than 16,000 candidate SNPs in the mouse. Using pooled DNA from 92 unrelated individuals and MALDI-TOF mass spectrometry, we have now validated more than 8,000 human SNPs. We have integrated our SNP discovery efforts with those from other workers to give a comprehensive view of gene-based SNPs. We provide a browser that shows the location of polymorphisms in mRNA sequences, and indicates whether variants cause amino acid substitutions. Coding regions and conserved protein motifs from the Pfam database are also displayed in the browser. If a SNP alters an amino acid in a conserved protein domain, we assess how the amino acid substitution affects the fit of the protein to the motif model. SNPs are projected onto known three-dimensional structures, when appropriate, so it is possible to assess their role in intermolecular interactions. The integrated maps, a Java-based tool for viewing candidate SNPs in the context of EST assemblies, reagent information (including PCR primers and extension primers), and a SNP search engine are available at our website: http://lpgws.nci.nih.gov/GAI/. We provide access to our SNP detection software for non-commercial use. We also have developed web-based tools for accessing the Quantitative PCR Primer Database maintained by the Gene Expression Laboratory at NCI-Frederick. Through our search engine, http://lpgws.nci.nih.gov/cgi-bin/PrimerViewer, researchers can retrieve information about PCR-based reagents for measuring the expression of human and mouse genes. The database currently includes 1938 human and 290 mouse primer sets. In addition, the GAI provides annotation for the Affymetrix U95 and U133 human, and M74 and MOE430 mouse GeneChip expression array sets. Oligonucleotide sequences from the expression arrays have been mapped against the complete set of mRNA and EST sequences in the current UniGene build. This data enables users to determine the current name and description of a gene assayed by a microarray probe set, visualize the position of probes within a transcript and verify whether probe is specific for its target. Annotation tools are available at http://lpgws.nci.nih.gov/cgi-bin/AffyViewer.cgi.